Venture Capital Sees Big Returns in Big Data

When Jason Goldberg set out to raise a new round of funding for his flash sales site Fab.com, he dispensed with the usual PowerPoint presentations and instead gave potential investors a look at the crown jewels: the "dashboard" of real-time analytics that can instantly spot trends and enable the site to tweak its offerings on the fly.  The company soon closed a $40 million round of funding from blue-chip venture capitalists including Andreessen Horowitz. 

 Such is the allure of big data, a buzzword that refers to the new found ability to collect and analyze massive amounts of information on almost every dimension of the human experience. In sectors ranging from retail to healthcare to social media, the promise of Big Data has venture capitalists salivating. "The best themes are the ones you can see everywhere and big data is everywhere you turn," said David Hornik, a partner at August Capital. 

Venture firms invested a total of $2.47 billion last year in fields around big data, including database management and data processing, according to Thomson Reuters data. That compares to $1.53 billion in 2010 and $1.1 billion in 2009. Venture investors are building their strategies around a few key themes within big data. Accel Partners' Ping Li controls a dedicated $100 million Big Data fund, carved out of bigger Accel funds. 

He divides the themes into four, starting with the storage and networking services that support big data platforms. Amazon [AMZN  182.50    2.57  (+1.43%)   ] <http://data.cnbc.com/quotes/amzn> , with its web-services business, is an example. 

The next group is the type of company that builds the platforms that enable the analysis of huge volumes of data. That would include his own portfolio company, Cloudera, and others such as Hortonworks, backed by Benchmark Capital and Yahoo [YHOO  15.01    -0.345  (-2.25%)   ] <http://data.cnbc.com/quotes/yhoo> . Cloudera and Hortonworks help companies work with Apache Hadoop, an open-source software framework that helps organize and manipulate big sets of data. 

Once the data is set up in an analysis-friendly way, businesses need software applications that make sense of it. Here, Li sees two more areas for investment — what he calls big data apps, and data-driven apps. A prominent big-data app is Domo, the business-intelligence service run by Omniture founder Josh James. It has snagged $63 million in funding from venture firms like Institutional Venture Partners and Benchmark Capital. Data-driven apps — meaning the data is used in a way that drives another app, rather than analyzed and managed for its own intrinsic value — range from companies like the Climate Corporation, which takes myriad data points about the weather and uses them to price crop insurance, to Lookout, which uses data to protect mobile phones. 

Others, such as John Connors of Ignition Partners or Jim Smith of Mohr Davidow, prefer to think of the apps category as consumer-focused or enterprise-focused. In the enterprise category, Connors gives the example of his and August Capital's portfolio company Splunk, which allows companies to analyze data in a way that in the past might have required expensive data warehouses and specialized, hard-to-deploy technology. Splunk filed for an initial public offering last month. "People can begin, in corporate situations, to use corporate data the way we as consumers use the internet," Connors said. He pointed to a phone company cross-referencing a customer's data use compared to her payment plan compared to a plan the phone company is pushing, which he said became a simple exercise on Splunk compared to a too-expensive and too-complicated process even just a few years ago. 

"The cost to build and maintain the data warehouse under old technology would be in the range of $10 million to $15 million, and take about two years," he said. On Splunk, it cost $1.5 million and took three months. Without the lower pricetag and simplicity of use, the customer nuances wouldn't have come to light, perhaps costing the company lost customers and revenue.  

Liquid Robotics, a VantagePoint Capital-backed company that deploys floating robots to measure vast amounts of ocean data, plans to start selling information about minute variations in ocean currents that wouldn't have been practical to track using old technology, said chief technology officer James Gosling. Fishing fleets could use the information to find areas with the best catch, he said, or shipping companies could use it to find the fastest routes. Consumer plays include companies like Google, Facebook, and Zynga. "Social media is not monetizable without big data," said Mohr Davidow's Smith, who believes it would be impossible to target ads precisely without it. Connors sees medical services and traditional retail as areas that are ripe for more big-data based investments. Boston-based Foundation Medicine crunches through massive data sets on patients' tumors to compare aberrations, treatments and outcomes, with the goal of refining cancer care and improving the quality and length of life for those with the disease. Its backers, including Google Ventures, Kleiner Perkins Caufield & Byers, and Third Rock Ventures, hope it will one day become a big business. Co-founder and Third Rock Partner Alexis Borisy says Foundation wouldn't exist without the ability to collect vast amounts of data on individual tumors and analyze it quickly and inexpensively. 

In the retail sector, Fab.com aggregates data including users' purchase history, membership date, purchase rates and total spending on a minute by minute basis. While some quibble that data sets in retail may not reach the massive level of terabytes or more that some think of as big data, others say the definition depends more on the expansive way the data is drawn together and used. In Fab's case, the data sets are connected to a software platform that digests them in real time and spots trends that help drive business decisions. Another method called cohort analysis lets Fab.com use data to monitor groups of users by their join date and compare their engagement levels, purchase rates and lifetime value to the site. "We look at RJ Metrics data on cohorts, on sources where they joined Fab, what they go on to do and then we make adjustments on our advertising targets," Goldberg said, referring to the company that provides its big-data software. 

Venture capitalists say the big-data wave is just starting to build. Li, for one, sees limitless opportunity in mobile. "The mobile device is the single best data-capture device ever," he said. "Always with you, and it generates a ton of data."